http://www.raphkoster.com/2007/03/29/etech07-the-core-of-fun/
It's a bit abstract really (and I find his idea of a "theory of fun"
somewhat oxymoronic) but it is good to think about.
In response to Andrew & Mitch's earlier posts:
Re metrics: I think Raph was talking about metrics of user enjoyment,
not user contributions; subtly different but let's assume they're the
same for now (contributions are clearly more measurable).
Having a metric of contribution is important. However, as the original
wiki crowd (the XP guys) put it: you have to "do the simplest thing that
could possibly work". To my mind that means avoiding unnecessary coding;
and where coding is unavoidable, you avoid unnecessary research into
user interface design.
For example:- TWiki and (perhaps?) MediaWiki already have built-in
metrics for who has posted the most stuff to the wiki. What they don't
have is a ratings system, but there are lots of ratings systems out
there. I really, really don't want to get sidetracked into experimenting
with developing the next version of Slashdot's "karma" or Kos's "mojo"
or Advogato's "Trust Metric", or whatever...[*]
Where I think this kind of "game design" mentality can help is in
thinking about the fundamental steps of using the browser: locating a
gene, pulling up the external links (GO, Flybase, PDB...), looking at
neighboring genes, looking at other nearby features on the chromosome,
etc. (compare this with the slide from Raph's talk listing all the
things that someone does when they go to Amazon)
I don't quite know what we do with this list of steps yet, and I don't
mean to imply that we should do frivolous things. The smooth scrolling
interface (that we started with) may superficially seem frivolous but
it's not: it directly benefits the "look at neighboring genes" and "look
at nearby features" steps. I agree that Mitch's panning animation might
possibly help with a sense of place too (BTW Mitch, I think that a
quadratic (x^2) function for the pan would probably physics-y in the
sense of Newton's 2nd law....)
The steps we have not yet really delved into are the external links,
especially things like ontology/browser mashups (which could be really
fun), and also the wiki upload interface. We haven't yet got a working
upload, but when we do, the user should DEFINITELY feel rewarded for
having uploaded something.
NB (at risk of repeating myself) I don't mean we should add frivolous
bells & whistles, just that this concept of reward is probably a good
one to hold in mind when developing & evaluating an interface.
Of course, this may all be a little premature, since we don't have the
basics yet, but it's also good to think about.
[*] OK, as I re-read this, I realize a lot of those other trust metrics
are secret... Actually, I think that if we allow "hotlists" of some kind
(i.e. ability to aggregate other peoples' annotations and offer them up
as a preset list), then the need for a trust metric might go away, or
the metric might emerge naturally from the hotlist somehow.
|
